Characteristics of persons who complied with and failed to comply with annual ivermectin treatment

Authors


Corresponding Author Joseph C. Okeibunor, University of Nigeria, Nsukka, Enugu 410002, Nigeria. E-mail: jokeibunor@yahoo.com

Abstract

Objective  To assess individual compliance with annual ivermectin treatment in onchocerciasis-endemic villages.

Methods  Multi-site study in eight APOC-sponsored projects in Cameroon, Nigeria and Uganda to identify the socio-demographic correlates of compliance with ivermectin treatment. A structured questionnaire was administered on 2305 persons aged 10 years and above. Two categories of respondents were purposively selected to obtain both high and low compliers: people who took ivermectin 6–8 times and 0–2 times previously. Simple descriptive statistics were employed in characterizing the respondents into high and low compliers, while some socio-demographic and key perceptual factors were employed in regression models constructed to explain levels of compliance among the respondents.

Results  Some demographic and perceptual factors associated with compliance were identified. Compliance was more common among men (54.4%) (P < 0.001). Adults (54.6%) had greater rates of high compliance (P < 0.001. The mean age of high compliers (41.5 years) was significantly older (35.8 years) (t = 8.46, P < 0.001). Perception of onchocerciasis and effectiveness of ivermectin influenced compliance. 81.4% of respondents saw benefits in annual ivermectin treatment, high compliance among those who saw benefits was 59.3% compared to 13.3% of those who did not (P < 0.001).

Conclusion  Efforts to increase compliance with ivermectin treatment should focus on providing health education to youth and women. Health education should also highlight the benefits of taking ivermectin.

Abstract

Objectif:  Evaluer l’adhésion des individus au traitement annuel à l’ivermectine dans les villages endémiques pour l’onchocercose.

Méthodes:  Etude multi-sites dans 8 projets parrainés par APOC au Cameroun, au Nigeria et en Ouganda afin d’identifier les corrélats sociodémographiques de l’adhésion au traitement à l’ivermectine. Un questionnaire structuré a été administréà 2305 personnes âgées de 10 ans et plus. Deux catégories de répondants ont été choisies afin d’obtenir à la fois des adhésions fortes et faibles: personnes ayant pris l’ivermectine 6 à 8 fois et celles l’ayant pris 0 à 2 fois auparavant. Des statistiques descriptives simples ont été utilisées pour caractériser les répondants dans les groupes d’adhésion forte et faible tandis que certains facteurs sociodémographiques et des facteurs clé de perception ont été employés dans les modèles de régression construits pour expliquer les niveaux d’adhésions parmi les répondants.

Résultats:  Des facteurs démographiques et de perception associés à l’adhésion ont été identifiés. L’adhésion était plus fréquente chez les hommes (54,4%) (p < 0,001). Les adultes (54,6%) avaient des taux plus élevés d’adhésion forte (p < 0,001. L’âge moyen des adhésions fortes (41,5 ans) était significativement plus élevé (35,8 ans) (t = 8,46, p < 0,001). Les perceptions sur l’onchocercose et sur l’efficacité de l’ivermectine influençaient l’adhésion. 81,4% des répondants voyaient les avantages du traitement annuel à l’ivermectine, l’adhésion forte parmi ceux qui percevaient les avantages était de 59,3% comparée à 13,3% parmi ceux qui ne les percevaient pas (p < 0,001).

Conclusion:  Les efforts visant à accroître l’adhésion au traitement à l’ivermectine devraient se concentrer sur la fourniture d’éducation à la santé aux jeunes et aux femmes. L’éducation sur la santé devrait également mettre en évidence les avantages de la prise d’ivermectine.

Abstract

Objetivo:  Evaluar el cumplimiento individual del tratamiento anual con ivermectina en poblados endémicos para oncocercosis.

Métodos:  Estudio multicéntrico de 8 proyectos financiados por APOC en Camerún, Nigeria y Uganda, para identificar los factores socioeconómicos correlativos al cumplimiento del tratamiento con ivermectina. Se administró un cuestionario estructurado a 2305 personas mayores de 10 años. Se seleccionaron dos categorías de respondedores para obtener tanto a los que más como a los que menos cumplían: aquellas personas que habían tomado previamente ivermectina entre 6–8 veces y entre 0–2 veces. Se utilizaron estadísticas descriptivas simples para caracterizar los respondedores que más y que menos cumplían, mientras que se utilizaron algunos factores socio-demográficos y de percepción claves en modelos de regresión construidos para explicar los niveles de cumplimiento entre quienes respondieron.

Resultados:  Se identificaron algunos factores demográficos y de percepción asociados con el cumplimiento. El cumplir era más común entre los hombres (54.4%) (p < 0.001). Los adultos (54.6%) tenían mayores tasas de cumplimiento (p < 0.001). La edad media de los más cumplidores (41.5 años) era significativamente mayor (35.8 años) (t = 8.46, p < 0.001). La percepción de la oncocercosis y la efectividad de la ivermectina influían en el cumplimiento. Un 81.4% de quienes respondían veían beneficios en el tratamiento anual con ivermectina, un alto nivel de cumplimiento entre aquellos que veían beneficios era del 59.3% comparado con un 13.3% de aquellos que no veían beneficios (p < 0.001).

Conclusión:  Los esfuerzos para aumentar el nivel de cumplimiento del tratamiento con ivermectina deberían centrarse en ofrecer educación sanitaria a los más jóvenes y a las mujeres. La educación sanitaria debería también hacer énfasis en los beneficios de tomar la ivermectina.

Introduction

The community-directed treatment with ivermectin (CDTI) programme of the African Programme for Onchocerciasis Controls (APOC) was established in 1995 (Amazigo & Boatin 2006) with the goal of maintaining a minimum of 65% annual population coverage in endemic communities over a minimum of 15 years for effective control of onchocerciasis (Plaisier et al. 1997; Borsboom et al. 2003; Tielsch & Beeche 2004). CDTI has been established in over 95 000 communities where more than 98 million ivermectin tablets are distributed annually to treat 33 million people (Amazigo et al. 2007).

Community-directed treatment with ivermectin is a process in which the community itself has responsibility for organisation and execution of treatment of its members (Remme 1997; Brieger et al. 2002). CDTI entails the empowerment of communities to decide on dates of distribution, mode of distribution (e.g. house-to-house, central place) and the persons who will guide distribution, the Community-Directed Distributors (CDDs). Other community responsibilities include conducting a census, collecting drug supplies, mobilisation during the distribution, and recording and reporting coverage. There may be one or more CDDs per community and playing a major role in conducting these activities (APOC 2007).

While reports of coverage are encouraging (APOC 2004; Amazigo et al. 2007), coverage rates in a community may not give the full picture of programme success because there may be individuals or groups who systematically do not comply over the years and thus provide a continued focus for transmission. Compliance, also known as adherence, is defined as ‘the extent to which a patient acts in accordance with the prescribed interval and dose of and dosing regime (IPSOR Medication Compliance and Persistence Special Interest Group 2006). Compliance with annual ivermectin treatment, therefore, has become a major challenge for APOC as it enters its second decade of implementation on the ground.

To date, very few localised published reports of CDTI have actually determined the extent to which individual community residents comply with ivermectin treatment consistently each year and their reasons for doing so or not. This may be due in part to the fact that in the early years of the programme not enough annual distributions had accumulated to provide a meaningful measure of compliance. Now that the original 25 projects, which started in 1997–1998, have been operating for almost a decade; annual compliance studies become possible. Such studies also become extremely desirable because researchers are now pushing back the timeframe for controlling onchocerciasis through annual ivermectin dosing from 15 to 25 or more years (Winnen et al. 2002). APOC has a mandate to establish within a period of 12 to 15 years, effective and sustainable, CDTI throughout the endemic areas within the geographical scope of the programme (APOC 2006) and thus requires a clear understanding of the long-term compliance process to guide countries toward sustainability.

Considering that ivermectin treatment will be needed for many years into the future, the Technical Consultative Committee of APOC requested a study to learn what factors might be associated with compliance over time so that appropriate education and intervention could be designed to help sustain annual treatment (APOC 2005). High levels of compliance are even more important now than at the time the study was conducted because of the shift in the objectives of APOC from ‘controlling onchocerciasis as a public health problem’ to ‘interrupting transmission where feasible’. Some studies have identified factors that encourage or discourage taking ivermectin during a given distribution such as age, gender and ethnicity (Brieger et al. 2002; Maduka et al. 2004; Semiyaga et al. 2005; Lakwo & Gasarasi 2006). Social support and drug perceptions are other factors that have influenced coverage, and hence may impact on compliance (Nuwaha et al. 2005). Akogun et al. (2000) discuss the importance of perceived benefits of ivermectin treatment, which could be another motivating factor in annual compliance. These factors may or may not explain long-term compliance. APOC therefore wants a scientific basis to frame health education and communication that will promote the sustainability of CDTI.

Compliance studies on other medical conditions have shown that adhering to a medical regimen may be influenced by the characteristics of the patient/client and provider as well as the nature of the regimen and may guide thinking about compliance with ivermectin treatment. Patient factors often include ethnic origin (Yuan et al. 2006). Educational level has a positive association with compliance (Iliyasu et al. 2005).

This study presents results of a multi-site study of compliance with ivermectin treatment in three APOC-assisted countries with CDTI programmes. The main aim was to identify the factors associated with levels of compliance among villagers with varying socio-demographic and perceptual characteristics. The conclusion of the study would help in designing programmes to ensure compliance to ivermectin treatment. Results are expected to contribute to developing a simple and efficient protocol for determining compliance rate at regular intervals, which could be used by the National Onchocerciasis Taskforces for periodic monitoring of compliance to treatment with ivermectin in their respective project areas.

Methods

Study design

The study was designed to allow an assessment of compliance to and perceived benefits of annual ivermectin treatment and the feasibility of assessing long-term compliance. The cross-sectional approach was adopted in collecting qualitative data from seven study sites in three countries that scaled through the feasibility study conducted in 2005.

The study took place in Cameroon, Nigeria and Uganda. These are the countries with CDTI projects that have been implementing CDTI since at least 1998 and where the feasibility study demonstrated adequate records at all levels. The levels in CDTI implementation are the Project (Provincial/State), District/Local Government Areas, Frontline and Community. Records of CDTI implementation are kept at each level and the records include, among others, the number of people treated each year as well as the number of distributions. These project areas include Cameroon (SW1 CDTI project), Nigeria (Kaduna, Imo, Taraba and Cross River States CDTI projects) and Uganda (Phase 1 and 2) CDTI projects.

Cameroon study site

The South West 1 (SW1) Project covers part of the south-west province of Cameroon and includes three administrative divisions (Fako, Kupe and Manengouba, Meme). SW1 is divided into eight health districts (Buea, Tiko, Bangem, Tombel, Nguti, Limbe, Muyuka and Kumba), which are subdivided into 61 health areas. The total area of SW1 is approximately 14 300 km2, and the population is estimated to 710 050 people living in 465 communities. The vegetation here is predominantly the equatorial rain forest. Beside this main type of vegetation, there is mangrove vegetation on the coast (Atlantic Ocean). The altitude ranges from 0 m on the coast to 2200 m in Buea town (headquarter of the province), with a multitude of small hills.

Mount Cameroon, the highest peak in West Africa with 4100 m of altitude, stands in Buea town. The SW1 has a very rich network of drainage systems most of which flow from high altitudes and are interrupted by numerous cascades, rapids and waterfalls constituting suitable breeding site for the simulium vector. The climate is tropical. The rainy season lasts from mid-March to mid-October with its peak around July and August. Farming is the main activity of population year-round. The rich volcanic soil encourages the cultivation of various food crops.

Distinct ethno-linguistic groups live in SW1. These include the Bakweries, Bafaws, Bakossis, Orokos and Balongs. The communities in the SW1 are generally permanent settlers. The official language of communication is English.

Nigeria study sites

With a population of more than 140 million in 36 States and a Federal Capital Territory (FCT), Nigeria has a total of 27 CDTI projects in 32 States, including the FCT, located in four health zones, namely north–west, north–east, south–west and south–east. These health zones serve as operational zones for APOC CDTI implementation in the country. The health zones are culturally distinct with varying health-seeking behaviour.

Ten of the CDTI projects, with a population of <9 million and an ultimate treatment goal (UTG) of <8 million in 16 003 communities covering the four health/CDTI operational zones, have been implementing CDTI since 1998. These projects are facilitated by the State Ministries of Health (SMOH), with financial and technical support of 5 NGDOs: SSI, CBM, MITOSATH, GRBP and UNICEF. Four projects, namely Cross River, Imo, Kaduna and Taraba, were selected for the study in Nigeria.

Imo State is located in the south-eastern region of Nigeria. It has an estimated total population of about 3.4 million persons, with 27 Local Government Areas, of which 16 LGAs are implementing APOC supported CDTI strategy for the control of Onchocerciasis. An estimated >1.1 million persons are at risk of onchocerciasis in these 16 LGAs. The supporting NGDO for the implementation of CDTI in these 16 LGAs is the Carter Centre. The major rivers are the Imo, the Otamiri, the Ogochia, the Nbaa and the Uramiriukwu. The climate of the State consists of the dry season (November–March) and the wet season (April–October). The main occupations of the people are farming, fishing and trading. Community leaders in consultation with the elders and community members make decisions.

Mectizan® distribution started in Imo State in 1993 in four LGAs – Okigwe, Ahiazu, Ihitte-Uboma and Ezinihitte LGAs. A total of 114 157 persons were treated in these LGAs by then. In 1998, a proposal made to APOC for Onchocerciasis Control in Imo and Abia States was approved, with Global 2000 Nigeria (now the Carter Centre) as technical partner.

The project is located in the south-eastern health zone, which corresponds to a culturally distinct area with a remarkable health-seeking behaviour. The people are predominantly Christians from the Ibo extraction. This is one of the three major ethnic groups in the Nigeria. Its inclusion also ensured data from a project area facilitated by one of the major NGDOs operating in Nigeria.

The Cross River CDTI project started the Ivermectin treatment programme initially with the community based ivermectin treatment (CBIT) strategy in 1995 in seven LGAs. By 1997, the African Programme for Onchocerciasis Control (APOC) approved funding for CDTI implementation starting in five LGAs. The project has succeeded in gorging an interesting network of collaboration with other indigenous NGOs and government organisation such as the Cross River National Park, Forestry Commission, Youthcare, south-eastern Nigeria Eye Care Services (SENECS) and Pace Setters who have been involved in promotion of CDTI implementation in the State. The Cross River CDTI covers a total of 14 LGAs with APOC funding and under the facilitation of UNICEF. It is one of the six States in south–south Nigeria Primary HealthCare (PHC) zone.

Kaduna State consists of 23 LGAs and a population of more than 5 million. The State shares boundaries with Katsina and Kano States in the north, Plateau State to the east and Nassarawa State to the south-east. Niger State borders it to the west and the FCT to the south. Some major rivers in the State include Kaduna and Gurara. These provide good breeding sites for the black fly vector. The vegetation is mainly savanna grassland with few areas of forest mosaic.

The REMO in Kaduna shows that 15 of the 23 LGAs are hyper-/meso-endemic for onchocerciasis, thus qualifying for mass treatment with ivermectin. APOC funding of the State CDTI for onchocerciasis control started in 1998. The commencement of implementation of CDTI in the 15 LGAs was phased out between 1998 and 2000. In 1998, five LGAs started and gradually by the year 2000, all the LGAs came on board.

Kaduna CDTI has implemented CDTI for over 7 years with the technical support from Sight Savers International. The project is situated in the north–western part of the country, with sizeable proportion of Moslems, with known gender ideology in the country, which may affect the participation of women in the mass treatment with ivermectin.

Distribution of ivermectin in Taraba State, using the CDTI approach started in July 1997. However, the project had distributed Mectizan® since 1994, using the mobile system and later the CBIT strategy. The number of people treated rose from a few thousands before the adoption of CDTI approach to over 880 000 persons by 2001 with the ultimate treatment goal (UTG) currently at 980 000.

The project consists of hyper-endemic foci with high prevalence of blindness. The project was initially supported by Africare and after its withdrawal by CBM and a local NGDO, MITOSATH. The International NGDO (CBM) mentored the local one (MITOSATH) and transferred responsibilities to MITOSATH within a given timeframe – as a way of promoting programme sustainability. This has resulted in the devolution of support for three LGAs by CBM to MITOSATH in 2001 and of the entire project area to MITOSATH in 2004.

The project is located in the north–east health zone of Nigeria. This is also the only project area, among those that have implemented CDTI for at least 7 years, that is supported by a local NGDO. It was expected that the compliance to and perceived benefit of annual treatment with ivermectin could be influenced by the activities of the supporting NGDOs. Thus, the inclusion of this project did not only provide data from a distinct health zone in Nigeria but also provided data from a project area that is supported by a local NGDO in Nigeria.

Uganda study sites

Phase I CDTI project.

Onchocerciasis is endemic in more than a third (22) of the Districts in Uganda, and about 1.46 million people are already infected with Onchocercal Volvolus. Approximately 1.34 million people are on treatment through mass distribution of ivermectin. Although onchocerciasis does not feature among the ten most burdensome diseases in most districts, its impact on quality of life of its victims has been found to be bigger than expected.

The implementation of CDTI in Uganda has been phased. The Phase1 Uganda CDTI project covers four districts, namely Kisoro, Kasese, Masindi and Hoima. These were the first areas to be mapped for onchocerciasis and had partners who were already involved in control of onchocerciasis. The NGDO partners supporting the Districts include Global 2000 and Sight Savers International. Two districts, Hoima and Kasese were covered in this study.

Phase II CDTI project.

Phase II CDTI project in Uganda includes Kabale, Bushenyi, Kamwenge, Kyenjojo, Kabarole, Sironko and Mbale districts. Kabale, Bushenyi, Mbale and Sironko districts are highland areas with an altitude between 1219 and 4321 m above the sea level. Mbale and Sironko have the highest altitude. Mbale and Sironko districts are located in Eastern Uganda, Kabale in south-western Uganda; Bushenyi, Kamwenge, Kyenjojo and Kabarole in Western Uganda. Kyenjojo, Kabarole and Kamwenge districts are areas of ethnic diversity including Batoro who are the indigenous people, and Bakiga and Bafumbira who have migrated from Kabale and Kisoro districts, respectively. Bushenyi is also an area of ethnic diversity, including Banyakole, who are indigenous people, and Bakiga and Bafumbira. Kabale, Mbale and Sironko districts have homogeneous populations with Bakiga in Kabale and Bagisu in Mbale and Sironko districts. The main economic activity in all phase II districts is agriculture with emphasis on food crops such as sweet potatoes, maize, cassava, beans, groundnuts and bananas. The cash crops are mainly coffee and tea.

Ivermectin distribution in Kabale, Mbale and Sironko districts started in 1993/1994 with assistance from the River Blindness Foundation, now called GRBP. Ivermectin distribution in Bushenyi (with assistance from CBM) and Kabarole district is now divided into three districts (Kamwenge, Kyenjojo and Kabarole) also started in 1991 with assistance from GTZ. The method of drug distribution was community based in all phase II districts except in Bushenyi district, where it was by mobile teams until 1995 when community-based methods were introduced. Community-directed treatment was introduced in all the 1409 communities in phase II districts in 1998 with financial support from APOC. Community here is defined as the smallest administrative unit headed by an elected village leader by community members. The study covered Kamwenge and Kyenjojo districts.

Study population

The study population consisted of people living in villages where ivermectin distribution has occurred using the CDTI approach since 1998, that is, eight continuous years. To be included in the study such people would have had to be at least 5 years old in 1998 so that they would have been eligible to receive ivermectin for each of the consecutive years. In addition, potential respondents were divided into two groups, high and low compliers. The former included those who were recorded as having taken ivermectin 6–8 times in the previous 8 years, while low compliers took ivermectin only 0–2 times.

Sampling

At the workshop in March 2005, the group found useful information on compliance and coverage in several published studies that helped to calculate the sample size using the sampling statistics. Two sample sizes were derived from computations using data from the Onchocerciasis Control Programme (OCP) and from publications on Nigeria and Uganda (See Emukah et al. 2004; Ndyomugyenyi 2004; Plaisier et al. 1997). However, the published data on Nigeria and Uganda were employed because they cover two of the countries included in this study. In the subsequent year, information was generated by our own feasibility study, which provided realistic data from the field (Brieger et al. 2007). With an estimated 62.3% compliance rate gathered from the feasibility study, and 95% confidence interval and 2.5% precision level at two tails, the upper limit of approximately 1500 individual household members was estimated per study site.

Two districts were randomly chosen from each project area, and 10 communities were randomly selected from each district. Using the community treatment registers in each of the sampled communities, 15 households were systematically selected. A sample of 20 individuals per community underwent more detailed interviews. These were stratified by the level of compliance and gender. Those taking ivermectin 5 or more times are termed ‘high compliers’, while those taking ivermectin 2 or fewer times are termed ‘low compliers’. On the whole, a total of 2320 respondents were approached in the study and all consented to participate. However, in the process of editing the final data, some were dropped for inconsistencies.

At each site, the research team balloted for 2 districts among those who were found to have conducted ivermectin distribution since 1998. Further, within each of the two districts, a random sample of 10 villages was chosen. CDD records were used to identify high and low compliers in each village.

Study instrument

The individual questionnaire was structured questionnaire designed to provide data on respondents’ compliance to ivermectin treatment as well as the demographic and perceptual factors that influence compliance with ivermectin treatment in the communities. The questionnaire was standardised in a workshop involving all the scientists after pretesting in communities outside the study areas but with similar characteristics with the study communities in each project area. The respondents were asked their perceptions of the likelihood of getting infected as well as perception of effective medication for preventing the infection. They were also asked to express their opinions on alternative ways of preventing onchocerciasis. The questions included are as follows:

  •  What are the different treatments you know for onchocerciasis (River blindness)?
  •  Which of this do you think is the best? (pick only one)
  •  Have you ever taken ivermectin since they started distributing it in this village?
  •  What motivated you to take ivermectin the first time? (Tick as many as mentioned)
  •  Do you think that taking Ivermectin annually for several years has had benefits?
  •  If yes, can you mention these benefits?

People were asked their likelihood of having any of the symptoms of onchocerciasis or perceived susceptibility to four symptoms including skin rashes, blindness, swelling of the body and severe itching. This was without reference to ivermectin and the lesson from that is that more of those who perceive the likely of being infected complied more. A susceptibility scale was constructed awarding 0 points for not susceptible, 1 for uncertain, 2 for likely and 3 for very likely for each of the four symptoms. The scores ranged from 0 to 12 with a mean of 4.0 points. High compliers scored an average of 3.7 points, while the low compliers had a significantly higher mean of 4.3 (P < 0.001).

Instrument administration

Research assistants trained on the methods and objectives of the study administered this questionnaire on 2305 persons aged 10 years and above. Those who were 10 years old in 2006 when the study was conducted would have had at least 6 treatments between 2000 when they become eligible and 2007 inclusive. Two categories of respondents were purposively selected to obtain both high and low compliers, that is, people who took ivermectin between 6 and 8 times previously and 0–2 times, respectively. Both at the village and individual levels, potential respondents were informed that the researchers are from the National Onchocerciasis Control Programme trying to learn about their experiences with the ivermectin distribution programme since 1998, that is, the year when (special event) happened.

Data analysis

Data were entered and analysed with the EPI version 6. Simple descriptive statistics were employed in characterizing the respondents while some socio-demographic and key perceptual factors were employed in regression models constructed to explain the levels of compliance among the respondents. Before running the regression models, however, the compliers were cross-tabulated with the socio-demographic characteristics of the respondents. Various responses of their perception of the onchocerciasis and the effective remedies were also cross-tabulated with the high compliers.

Ethical considerations

Ethical clearance was obtained from ethical clearance committees of the University Teaching Hospitals in Enugu, Kaduna in Nigeria and Buea in Cameroon. Permissions were also sought from the community leaders. An informed verbal consent was obtained from every eligible individual before inclusion into the study by explaining the objective of the research. Privacy and confidentiality were ensured. After the interview, data collectors provided important information regarding onchocerciasis mainly focusing on misperceptions and knowledge deficits observed during the interview with the respondents.

Results

The individual questionnaire was completed for 2306 persons selected from the CDD register because of their high (6–8 times, 50.7%) or low (0–2 times, 49.3%) compliance with annual ivermectin treatment. Just under half (49.0%) of respondents were women. The majority (79.1%) were adults aged 25 years of age and older. The average age was 38.7 years. Over one-third (46.2%) were currently not married. One-third (33.9%) of respondents had no formal education, while 38.8% had primary schooling 27.2% had post-primary education. Nearly two-thirds (63.1%) listed their primary source of income as farming, hunting or fishing. Other occupations include trading (16.1%), civil service (10.8%) and artisanry (10.0%). Ethnic minorities living among villagers of the local majority ethnic group comprised 4.6% of the sample. The large majority (94.6%) were Christians, while the remaining adhered to Islam and other faiths.

Table 1 presents comparison between compliance and the demographic aspects of the individual questionnaire. High compliance was more common among men (54.4%) than women (46.9%) (P < 0.001). Adults (54.6%) had greater rates of high compliance than youth (36.0%) (P < 0.001). The mean age of high compliers (41.5 years) was significantly older than that of low compliers (35.8 years) (t = 8.46, P < 0.001).

Table 1.   Demographic factors associated with individual compliance
FactorNumber% High Compliance ≥5 treatmentsχ2 (Yates for 2 × 2) P-value
Sex
 Male117554.412.60
<0.001
 Female113046.9
Age
 Youth (10–24 years)48136.052.32
<0.001
 Adult182354.6
Current marital status
 Married123853.89.08
0.003
 Not106147.4
Religion
 Christian182451.50.59
0.44
 Non-Christians10447.1
Education
 None78254.724.95
<0.001
 Primary89553.1
 Post-primary62742.3
Ethnic
 Indigene of village219652.346.38
<0.001
 Minority in village10617.9
Occupatio
 Farmer/fishing/hunting145458.696.42
<0.001
 Non-Farmers/fishers/hunters85237.3

Those currently married (53.8%) were more likely to have high compliance than those not presently married (47.4%) (P = 0.003). There was a small difference in rates of high compliance between Christians (51.5%) and those of other faiths (47.1%) (P < 0.001). Education level for individual respondents was inversely associated with high compliance: 54.7% with no education, 53.1% with primary schooling and 42.3% with at least secondary education. Farmers, hunters and fishermen (58.6%) had a larger proportion of high compliance than those of other occupations (37.3%).

Ethnic variation was seen. Among people of the majority population in a particular village, 54.8% had high compliance, 52.3% of ethnic minorities living in their own villages were high compliers, but only 17.9% of ethnic minorities living among the majority had high compliance (P < 0.001).

Perceptual factors influencing compliance

A variety of items on the individual questionnaire addressed possible perceptual issues including perceptions of benefits and problems of ivermectin treatment as well as perceived seriousness and symptoms of onchocerciasis (Table 2). A total of 1877 (81.4%) respondents saw benefits in annual ivermectin treatment, high compliance among those who saw benefits was 59.3% compared to 13.3% of those who did not mention benefits. Both social and health benefits were mentioned, but health was the most common and included improved sight, deworming, reduced itching, improved skin and increased strength. The main social benefits mentioned were ability to work better and ability to work longer and harder.

Table 2.   Perceptual factors and ivermectin compliance
FactorResponseNumber% High Compliance ≥5 treatmentsYates corrected χ2, P-value
See benefitsYes187759.3293.91
<0.001
No42913.3
See problemsYes39628.891.10
<0.001
No191055.3
Best medicine for onchocerciasisOrthodox209853.152.93
<0.001
Indigenous/other20826.4
Perceived seriousnessYes209652.838.84
<0.001
No21030.0
Someone encouragesYes210952.426.36
<0.001
No19733.0
Someone discouragesYes31220.2133.28
<0.001
No199455.5
Most people takeYes211853.577.64
<0.001
No18819.7
Community willing to takeYes208753.045.76
<0.001
No21928.8
Perceived susceptibility scoreCompliance statusNumberMean SDt value
P-value
High11703.7, 3.94.31
<0.001
Low11364.3, 3.7

Problems in taking ivermectin were voiced by 396 (17.2%) people. The most common were side effects like itching, swelling and dizziness. Inconvenience was also mentioned and related to both experiencing side effects that might disrupt daily lives and work as well as the inconvenience of meeting up with the distribution. Only 28.8% of those who saw problems had high compliance compared to 55.3% who mentioned no problems.

Most (91.0%) thought western orthodox medicines including ivermectin and banocide were best for treating onchocerciasis. Among these, 53.1% had compliance compared to 26.4% who thought indigenous medicine and other forms of treatment were better (P < 0.001). Comparison of perceptual factors and compliance is shown in Table 3.

Table 3.   Key demographic and perceptual factors associated with compliance
SourcedfSum of squaresMean squareF-statistic  
  1. Age refers to the age of the respondents; Ethnic refers to whether the respondent is a minority ethnic group or not; Mosttake refers to perceptions of most people taking ivermectin in the communities; seebene refers to whether the respondents perceive any benefits in taking; seeprob refers to those who perceive problems with taking ivermectin; bestmed refers to perception of ivermectin as the best treatment for onchocerciasis.

Correlation coefficient: r2 = 0.24, ra2 = 0.24
Regression9138.047315.338680.42  
Residuals2289436.59820.1907   
Total2298574.6455    
 B95% confidencePartial
VariableMeanCoefficientLowerUpperSEF-test
B coefficients
Sex0.50890.04847140.0124130.0845300.0183886.9490
Age38.69030.00334290.0022140.0044720.00057633.7304
Ethnic0.95390.28198630.1946160.3693560.04455340.0594
Farmer0.62980.13790100.0992650.1765370.01970248.9920
Mosttake0.91870.11170920.0419590.1814590.0355689.8641
Discour0.1353−0.2135586−0.270204−0.1569130.02888554.6609
Seebene0.81380.32172820.2713720.3720840.025678156.9831
Seeprob0.1718−0.1767417−0.227311−0.1261720.02578746.9755
Bestmed0.90950.11657850.0512540.1819030.03331112.2479
Y-intercept−0.4143358     

Most respondents (90.9%) believed that onchocerciasis was a serious disease. Of these 52.8% were high compliers. Among those who did not rate onchocerciasis as serious, only 30.0% were high compliers (P < 0.001). Over 90% of respondents said someone had encouraged them to take ivermectin. Among those people, 52.4% had high compliance, while only 30.0% of those who reported no encouragement were high compliers (P < 0.001). Actions taken to encourage people included someone telling them of the benefits, house-to-house visits, community mobilisation and passing information when the drug is available.

Few respondents said that anyone actually discouraged their taking of ivermectin (13.5%). Among those who had experienced discouragement, 20.2% were high compliers. In contract, 55.5% of those who reported no discouragement had high compliance (P < 0.001). Forms of discouragement included hearing someone express fear of side effects and being told not to take it because of sickness. A few were told that ivermectin was not for humans and that the drug had expired.

Two questions addressed perceptions of what others in the community do and think. When asked whether most people in the community take ivermectin, 91.8% agreed. Among those who thought most take the medicine, 53.5% were high compliers compared to 19.7% of low compliers. Concerning perceptions of whether the community was willing to continue ivermectin distribution, 90.5% thought they were willing. Among these 50.3% were high compliers compared to only 28.8% of the low compliers (P < 0.001).

Further analysis was carried out combining the demographic and perceptual factors from the individual questionnaire. The resulting analysis found nine factors that were associated with the level of compliance (Table 3). Four were demographic: being men, adult (25 years and older), in the ethnic majority of their village and a farmer/hunter/fisherman. Three perceptual factors showed a positive association with high compliance and included believing most people take ivermectin, seeing benefits in annual treatment and thinking western medicine was the best for managing onchocerciasis. The two perceptual factors that were negatively associated with compliance were seeing problems with ivermectin treatment and feeling that others were discouraging taking the medicine.

Discussion and conclusion

While reports of annual ivermectin distribution provide encouragement that villages as a whole are able to meet annual targets of 65% population coverage (Kennedy et al. 2002; Ndyomugyenyi & Kabatereine 2003), it was possible in this study to identify a fairly large pool of people who had taken in less than three times in the 8 years since the programme began in the study areas. These low compliers represent a potential reservoir of onchocerciasis infection that could thwart efforts to control the disease in accordance with observations by Borsboom et al. (2003) that, ‘When the parasite reservoir is not entirely eliminated, it is unclear whether ivermectin treatment can be stopped’. This is especially important as predictions on the required life of ivermectin treatment programmes vary up to 25 years or more (Winnen et al. 2002).

So far models and predictions of onchocerciasis control have been predicated on the assumption that individual compliance will be normally distributed (Winnen et al. 2002). This may not be a valid assumption in the context of the present study. The needs to understand the factors associated with systematic non-compliance among certain segments of the population led to the current effort to understand the characteristics and perceptions of both high and low compliers, which has identified factors that should prove useful in designing health education and behaviour change communication that will hopefully improve and maintain higher levels of compliance in the upcoming years.

Previous work has identified certain sub-groups with lower annual coverage (Brieger et al. 2002; Maduka et al. 2004; Semiyaga et al. 2005; Lakwo & Gasarasi 2006), and this study confirms that this challenge persists over the longer term. In 2005, a step-down process from project to district to community level was used to identify APOC project sites where continuous ivermectin distribution up through 2004 had occurred. The first step consisted of selecting 17 of 25 projects by APOC. The second step determined adequacy of districts where distribution had occurred on a regular annual basis. Among 121 districts, 58.6% undertook distribution in all 7 years. A total 852 villages were visited and community level records were found in all but three. Records showed that distribution had occurred for a minimum of five consecutive times in 429 villages, and ultimately 10 projects. The findings revealed that while the feasibility study found an adequate number of villages to study compliance, the large number of projects, districts and villages that did not qualify for the follow-on compliance study should lead National Onchocerciasis Control Programme managers to strengthen the overall coverage and consistency of their efforts (Brieger et al. 2007). It is not uncommon for minority ethnic groups to be left out of disease control programmes by being overlooked or ‘unseen’ by health workers and neighbours. (Dao & Brieger 1994/95; Brieger et al. 1997).

While health education cannot change demographic factors, it can help pinpoint certain groups who are most at risk for non-compliance. In particular, community distributors of ivermectin need to be taught during annual refresher training about the need to reach out to ethnic minorities, women and youth.

While non-farmers do not form a specific sub-group of the population, the association between compliance and farming/hunting/fishing implies that certain occupations are related to stability in the community and thus greater opportunity to comply. Thus, CDDs should be aware of people whose occupations take them away from the village and guide the village to plan distribution in such a way as to reach these people whenever they are at home. Strategies such as involving groups with lower compliance in village committees that plan the annual distribution could be adopted.

The perceptual factors offer several suggestions on how to train CDDs and front line health facility workers to communicate with the villages. Those items tested and confirmed in large part conform to the Health Belief Model (HBM) as tested against compliance by others. (Austin et al. 2002; Russell et al. 2006) HBM in simplest terms takes a two-pronged look at the likelihood of complying with a recommended health action. One set of factors consider threat perceptions of the disease in question consisting of perceived susceptibility and seriousness (severity). The second broad set looks at views of the proposed action itself and the perceived benefits and barriers (problems) inherent in adopting that action.

Although susceptibility has been associated with compliance in other studies (Russell et al. 2006), in the case of ivermectin compliance, perceived susceptibility did not make the final cut. This may not be unconnected with the belief in many communities that all people are born with some onchocerciasis in their bodies, and susceptibility to the consequences depends on how much onchocerciasis one has (Brieger et al. 1986) In this case, perceived severity or seriousness was important and has frequently been associated with compliance in other studies. (McDonald-Miszczak et al. 2004; Lange & Piette 2006; DiMatteo et al. 2007).

Other studies have variously documented the association of perceived benefits (Finney Rutten & Iannotti 2003; Farquharson et al. 2004) and barriers (Orensky & Holdford 2005; Russell et al. 2006) with taking a recommended action, in this case with compliance. Yirga et al. (2010) have also documented other perceptual factors that influence compliance with ivermectin treatment in Ethiopia. In terms of ivermectin compliance, perceptions of both benefits and problems/barriers were associated with the former showing a positive relation, and logically the latter being negatively associated. One may also include positive beliefs about orthodox medicines being best for managing onchocerciasis as a potential benefit. Although a higher proportion of villagers perceived benefits than problems, it is the latter that may contribute more to the continuing pool of infected people.

The issue of side effects is simple on the surface. These mainly occur during the first and possibly the second time that some people take ivermectin. (Rothova et al. 1989) Not everyone reacts. Unfortunately inadequate health education about potential side effects has been associated with refusal to take ivermectin (Haselow et al. 2003) One might interpret the presence of people who have complied only once or twice as possibly reflecting those who had side effects and were afraid to continue treatment. CDDs need to be equipped with a simple village drug kit that contains analgesics and antihistamines to treat side effects. They also need counselling skills to explain that once people become used to the medicine, side effects will no longer be a problem.

Although HBM does not have a specific social reinforcement component, it does provide the opportunity to examine encouragement and discouragement by others as important ‘cues to action’. These cues often come from health workers and community members. (Austin et al. 2002; Farquharson et al. 2004) Family members have also been singled out as important encouragers of compliance. (Lee & Molassiotis 2002; Turrisi et al. 2006) In the case of ivermectin, CDDs need continued refresher training that emphasises their role as encouragers. Community meetings need to be used to stress the importance of neighbours helping each other remember and take part in annual distribution. CDDs and community leaders also need to be alert to people who spread false rumours about ivermectin and nip these in the bud.

The use of CDTI treatment registers for sampling frame, which is often incomplete, appears to be a major limitation for this study. However, this did not affect the study as feasibility study to ascertain the availability of data for the study was undertaken and data sources confirmed before the study commenced. In terms of the validity of the instrument used, only the face validity, which was performed by the experts in the field, was conducted. All the same, the pretesting of the instrument help to ensure its validity.

In conclusion, this study has identified some of the social and perceptual characteristics of people who have not complied or performed so at consistently low levels over 8 years of annual ivermectin treatment. This is an important segment of the population of onchocerciasis-endemic areas and needs immediate attention if control efforts are going to seriously limit the reservoir of infection for this debilitating disease. Health education strategies have been suggested. These require full community participation, which is in keeping with the philosophy of APOC’s efforts, but they are also labour intensive. Advocacy will be needed to ensure that international, national and local commitment to control is sustained in adequate measure to implement compliance promotion activities even after APOC funding winds down.

The CDDs themselves need better education about the ‘biology’ that connects probability of disease symptoms, probability of side effects and ivermectin treatment/compliance with annual treatment so they can confer this knowledge to the population. In other words, the CDDs need to understand that it is not a matter of people ‘getting used to the medicine’ who results in lower probability/severity of side effects, but that it is the beneficial effect of the drug on the disease/parasite in 1 year that results in lower probability/severity of side effects in the next year.

Ancillary